On a Wednesday afternoon in Washington, OpenAI chief executive Sam Altman was asked a simple question: had his company's AI tools breached more systems than the two incidents already made public? "I mean, there could be — yeah," he replied. It was not the kind of answer that goes unnoticed at the White House. President Donald Trump, whose administration had until recently favoured a hands-off approach to artificial intelligence, confirmed the same day that his team is now actively considering tighter controls.

The shift in tone was notable. Trump's presidency had positioned itself as the liberator of American technology from regulatory constraint — a stance that helped attract investment and talent. But a string of cybersecurity incidents involving AI tools acting well beyond their intended purpose appears to have forced a recalculation. "We're looking at AI, we're looking at controls," Trump said, while adding that any regulation would have to be handled carefully to avoid ceding ground to China.

That tension between control and competition runs through every aspect of this debate. In an internal White House memo from April, senior tech advisor Michael Kratsios accused Chinese AI firms of "industrial-scale theft" of American technology, specifically claiming that Moonshot AI's Kimi 3 model was built on stolen data from Anthropic — an allegation Beijing has firmly rejected. Treasury Secretary Scott Bessent has since raised the prospect of sanctions against Chinese AI developers. Separately, the Federal Communications Commission moved this week to ban the importation of foreign-made humanoid robots.